AKA, what I’m telling my team about using AI at work
AI is not going away
Humanity loves shortcuts. Calculators, iPhones, GPS, Netflix, Waymo – they’re all tools we’ve invented to offload our thinking to machines. And our brains haven’t evolved much in the last 70,000 years, so we need the help.
AI is cognitive offloading on a new scale. It spans every type of cognitive task, it’s getting markedly better with every AI model release, and it’s insanely cheap (free). Plus, Big Tech wants us addicted to it – and they have the capital, compute, and PhDs to do it.
Suffice to say: Every knowledge worker will be impacted by AI in the very near future. Adoption will keep increasing (800 million ChatGPT users so far), AI will move up the org chart in its capabilities, and most teams will get smaller. And increasingly, AI will be built into every knowledge work tool we use – so choosing to abstain will be harder and harder (and is dumb anyway). Why would you not want to borrow 30 points of IQ?
The two choices for us, as workers and an organization
Assume for these reasons, AI’s ubiquity is inevitable. You have two choices.
Choice 1: Be an AI Freeloader. Lazily accept AI into your life as the cognitive version of Waymo. Ask ChatGPT to draft the strategy/code/copy, give it a brief once-over, and copy-paste into Word or Slack. Do this once or twice as a “shortcut” when you’re busy — then get used to doing it every day, for everything. For a short time, get praised for speed/efficiency (and better grammar, for a lot of us). Long-term, get laid off, since you’re not adding anything to the output.
Choice 2: Be an AI Manager. Use AI to go deeper and push past obvious insights. Don’t just use AI to draft documents – use it as a conversation partner to challenge your thinking. Command a team of AIs to do work that meets your high standards. Never let documents out the door without questioning their insights and making them smarter and more differentiated. (Remember, we all get the same AI superpower – so those that use it better will do better work.) Short-term, get told you could do it faster if you offloaded more. Long-term, become the most valuable member of the workforce.
This is even more important at the company level. Companies with “AI Freeloader” employees will have undifferentiated, uncompetitive products and marketing strategies. Companies with “AI Manager” employees will have real competitive advantages. But a lot of companies won’t be in that second group, because they’ll have (unconsciously) encouraged employees to cognitively offload the lazy way.
My goal as a leader is to push you to be in the second group. Unlike the CEOs at Shopify, Duolingo or Fiverr, I’m not just putting out a company mandate to use AI. I’m putting out a company mandate to manage your AI, not offload your thinking to it.
I do not want to see strategies, plans, or insights that haven’t been pressure-tested with AI. And I don’t want to see strategies that haven’t been given human review and pushback. If you don’t use AI OR you copy-paste with AI, you don’t belong here … and you’re risking your place in the human workforce.
My expectations for AI use as CEO
If you want to work at Section, here’s a short list of my expectations how you use AI:
- Use AI every day: Claude, ChatGPT, and Perplexity are our three current default AIs.
- Develop your own suite of favorite AI tools, and request reimbursement for them.
- Run every medium-to-high-stakes decision past AI, but DO NOT let AI make the decision (or say in meetings, “ChatGPT told me we should…”).
- Run workflow audits for yourself and your teams and look for opportunities to introduce AI into those workflows. We will show you how to do this.
- Share successful use cases. Do not hoard proprietary advantages you created with AI. This is unacceptable.
- Learn from each other and don’t duplicate innovation. Look at our Section GPT library before you build something new, and see if someone else has already solved it.
- Only hire people who are AI-proficient (and ideally AI-obsessed).
- Attend our AI Lunch and Learns. This is expected and part of your job.
- If you’re a leader, you MUST be a power user – working high (using AI as a strategic thought partner) and low (using AI to create repetitive outputs).
- Every 6 months, review the impact of AI on your team, and report back your planned changes to headcount.
In summary, do not be an AI Freeloader. Be an AI Manager. Manage your AIs so that your work aligns with business outcomes, not outputs.
What about jobs?
No CEO has the answer on what AI will do to jobs. Right now, I believe that our teams will be affected in three ways. Some teams will be bigger (the ones where AI meaningfully accelerates productivity and impact). Some will stay the same size (the ones where more can be done without adding headcount). And yes, some teams will be smaller (the ones where AI can replace or automate human work).
In individual roles, people who define their value by outcomes, not outputs will be most valuable – as well as people who can do more than one thing. If you’re a specialist whose specialty is AI’s superpower, you need to diversify.
AI will move some people up much faster. These will be people with the ambition and application to drive our business forward. They may not look like traditional leaders, but in this new world, they will be.
The good news for our team: Our business is growing. Jobs and tasks will change, but I don’t expect full-time roles to go away any time soon. But this could change. All I can promise is that I’ll update you on my thinking every six months, I’ll involve the leadership team in any decisions, and I’ll be transparent about the impact of AI on our business and headcount needs.
I hope this is helpful to you in writing your AI manifesto. If you draft one, I’d love to see it. Email me at greg@sectionai.com.
